Trust Center

How we keep your data safe — in plain English.

No jargon. Here is exactly what Frontier does to protect your business, and what we're still working on.

Frontier was built on one rule: tell the truth. That goes for your data too. Below is what we actually do — each point reflects how the system is built today, not a wish list. Where something isn't finished, we say so.

Last reviewed June 16, 2026

Keeping your data separate (security)

  • Your data is walled off from everyone else's. The database enforces this on every table that holds your records, and an automated test proves one team can never read another team's data. That test runs on every single code change — if isolation ever broke, the build would stop.
  • Two-factor sign-in (2FA). You can turn on app-based 2FA, and changes to sensitive personal data require it.
  • Your saved API keys are off-limits to the browser. Keys you connect (like Twilio or RentCast) can only be read by our secure server. The database flatly denies the browser any access to them.
  • Everything travels encrypted. All traffic uses HTTPS/TLS, enforced everywhere — no exceptions.

Your data is yours

  • Export anytime. You can download your team's data whenever you want.
  • Delete on request. We can erase a specific person's personal information when you ask — a real "right to be forgotten," not a promise on paper.
  • Clear retention. We keep your data while your account is active. After you cancel, you get 30 days to export; then it's permanently deleted, and backup copies are purged within 90 days.
  • We never sell your data. See our Privacy Policy for the full detail.

How we use AI

  • AI runs on our server, never in your browser. We use Anthropic's Claude for writing and analysis, and Deepgram for turning calls into text. Your keys and data are never exposed to the browser for these.
  • Recording is opt-in and consent-first. Before any call is recorded, the app makes you confirm consent and stamps the time. Many states — including Washington — require everyone on the call to agree. You must get that consent first.
  • We send the AI only what the task needs. We're tightening this continuously (see below).
  • Per Anthropic's published policy, your prompts are not used to train their models.

Outreach & compliance

  • Opt-outs are respected automatically. Texts handle STOP and HELP for you, and anyone marked Do-Not-Contact is blocked from messaging.
  • Read the Terms and Privacy Policy for the full picture.
  • Your outreach is your responsibility. Frontier gives you the guardrails, but following calling/texting law (like TCPA and carrier 10DLC registration) is yours to own.

Accessibility

We run an automated accessibility check on every code change to catch common issues (contrast, labels, keyboard use) before they ship.

Reliability

When something breaks, we're upfront about it. Check our live status page for current system status and any incident updates.

In progress

What we're still building

Being honest means naming what isn't done yet:

  • Formal data-processing agreements with our AI vendors, including written no-training terms, are being finalized.
  • We are not SOC 2 certified yet. We're building toward it and won't claim it until an auditor signs off.
  • Automated offsite backups with a regularly tested recovery plan are on our near-term roadmap.
Questions about security or your data? Email info@scsfrontier.com.